@inproceedings{tian-etal-2018-polarity,
    title = "Polarity and Intensity: the Two Aspects of Sentiment Analysis",
    author = "Tian, Leimin  and
      Lai, Catherine  and
      Moore, Johanna",
    editor = "Zadeh, Amir  and
      Liang, Paul Pu  and
      Morency, Louis-Philippe  and
      Poria, Soujanya  and
      Cambria, Erik  and
      Scherer, Stefan",
    booktitle = "Proceedings of Grand Challenge and Workshop on Human Multimodal Language (Challenge-{HML})",
    month = jul,
    year = "2018",
    address = "Melbourne, Australia",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/W18-3306/",
    doi = "10.18653/v1/W18-3306",
    pages = "40--47",
    abstract = "Current multimodal sentiment analysis frames sentiment score prediction as a general Machine Learning task. However, what the sentiment score actually represents has often been overlooked. As a measurement of opinions and affective states, a sentiment score generally consists of two aspects: polarity and intensity. We decompose sentiment scores into these two aspects and study how they are conveyed through individual modalities and combined multimodal models in a naturalistic monologue setting. In particular, we build unimodal and multimodal multi-task learning models with sentiment score prediction as the main task and polarity and/or intensity classification as the auxiliary tasks. Our experiments show that sentiment analysis benefits from multi-task learning, and individual modalities differ when conveying the polarity and intensity aspects of sentiment."
}